Application of multilevel confirmatory factor analysis to compositional organizational constructs
International Journal of Organization Theory & Behavior
ISSN: 1093-4537
Article publication date: 15 September 2022
Issue publication date: 10 November 2022
Abstract
Purpose
This study aims to offer insight into the applications of multilevel confirmatory factor analysis (MCFA) within organizational behavior (OB) studies, specifically when using macro-level variables. The authors used the case of a scale measuring collective organizational engagement (COE) to demonstrate the advantage of using MCFA over single-level confirmatory factor analysis (CFA).
Design/methodology/approach
The authors used nine referent-shift items to measure the dimensions of physical, cognitive and emotional COE. MCFA and aggregated single-level CFA were conducted on the same data using Mplus V. 8.7, and the authors compared standardized parameters from the two techniques.
Findings
The results indicate a three-factor model of collective engagement with CFA and MCFA. Stronger item loadings, factor correlations and composite reliability were found in the MCFA within-level model compared to aggregated CFA model. MCFA also supported an alternate one-factor-between-three-factors-within model of COE.
Originality/value
This study demonstrates the application of MCFA with a compositional construct and compares MCFA to mean-aggregated single-level CFA. It presents a case for OB researchers to apply MCFA to compositional constructs as a best methodological practice.
Keywords
Acknowledgements
Funding: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Citation
Ganesh, P. and Srivastava, K.B.L. (2022), "Application of multilevel confirmatory factor analysis to compositional organizational constructs", International Journal of Organization Theory & Behavior, Vol. 25 No. 3/4, pp. 204-220. https://doi.org/10.1108/IJOTB-04-2022-0065
Publisher
:Emerald Publishing Limited
Copyright © 2022, Emerald Publishing Limited